Diabetes mellitus is a major cause of blindness due to irreversible structural and functional changes to the lens and retinal vasculature. Mounting evidence from work with experimental animals suggests that aldose reductase, the first enzyme of the polyol pathway of glucose metabolism, plays a key role in the pathogenesis of diabetic eye disease. Inhibition of aldose reductase provides a therapeutically attractive means to delay the onset and/or progression of diabetic complications in the eye. However, effective drug therapy will depend on inhibitors with a high degree of binding specificity - a criterion not met by the currently available inhibitors. The aldo-keto reductase superfamily contains enzymes with structural and kinetic properties similar to aldose reductase. Many of these aldo-keto reductases are high affinity receptors for the same aldose reductase inhibitors previously evaluated but withdrawn from clinical trials. The long-range objective or our application is to identify structural features of aldose reductase and related enzymes that explain their functional differences and perhaps provide distinguishing features that can be exploited in drug design. A corollary goal is to establish the physiological role of aldose reductase, as long-term inhibitor therapy is likely to be required for effective prevention of diabetic eye disease. Three specific aims are proposed to address these goals: (1) We will characterize the functional properties and expression pattern in the normal and diabetic eye of a newly discovered human enzyme that appears to have functional properties strikingly similar to aldose reductase; (2) Using a combination of mutagenesis and x-ray crystallography, we will test the hypothesis that the C-terminal domain is a structural feature that distinguishes the function of aldose reductase from other closely-related enzymes; (3) We will test the hypothesis that the physiological role of aldose reductase has been conserved in the budding yeast S. cerevisiae and that strains containing deletions of yeast aldose reductase can be marker rescued by the human aldose reductase gene.